valr (version 0.5.0)

bed_intersect: Identify intersecting intervals.

Description

Report intersecting intervals from x and y tbls. Book-ended intervals have .overlap values of 0 in the output.

Usage

bed_intersect(x, ..., invert = FALSE, suffix = c(".x", ".y"))

Arguments

x

tbl_interval()

...

one or more (e.g. a list of) y tbl_interval()s

invert

report x intervals not in y

suffix

colname suffixes in output

Value

tbl_interval() with original columns from x and y suffixed with .x and .y, and a new .overlap column with the extent of overlap for the intersecting intervals.

If multiple y tbls are supplied, the .source contains variable names associated with each interval. All original columns from the y are suffixed with .y in the output.

If ... contains named inputs (i.e a = y, b = z or list(a = y, b = z)), then .source will contain supplied names (see examples).

Details

input tbls are grouped by chrom by default, and additional groups can be added using dplyr::group_by(). For example, grouping by strand will constrain analyses to the same strand. To compare opposing strands across two tbls, strands on the y tbl can first be inverted using flip_strands().

See Also

http://bedtools.readthedocs.org/en/latest/content/tools/intersect.html

Other multiple set operations: bed_closest, bed_coverage, bed_map, bed_subtract, bed_window

Examples

# NOT RUN {
x <- trbl_interval(
  ~chrom, ~start, ~end,
  'chr1', 25,      50,
  'chr1', 100,     125
)

y <- trbl_interval(
  ~chrom, ~start, ~end,
  'chr1', 30,     75
)

bed_glyph(bed_intersect(x, y))

bed_glyph(bed_intersect(x, y, invert = TRUE))

x <- trbl_interval(
  ~chrom, ~start, ~end,
  'chr1', 100,    500,
  'chr2', 200,    400,
  'chr2', 300,    500,
  'chr2', 800,    900
)

y <- trbl_interval(
  ~chrom, ~start, ~end, ~value,
  'chr1', 150,    400,  100,
  'chr1', 500,    550,  100,
  'chr2', 230,    430,  200,
  'chr2', 350,    430,  300
)

bed_intersect(x, y)

bed_intersect(x, y, invert = TRUE)

# start and end of each overlapping interval
res <- bed_intersect(x, y)
dplyr::mutate(res, start = pmax(start.x, start.y),
                   end = pmin(end.x, end.y))

z <- trbl_interval(
  ~chrom, ~start, ~end, ~value,
  'chr1', 150,    400,  100,
  'chr1', 500,    550,  100,
  'chr2', 230,    430,  200,
  'chr2', 750,    900,  400
)

bed_intersect(x, y, z)

bed_intersect(x, exons = y, introns = z)

# a list of tbl_intervals can also be passed
bed_intersect(x, list(exons = y, introns = z))

# }